Augusta: From RNA-Seq to gene regulatory networks and Boolean models

Area of Science:

• Computational biology
• Systems biology
• Bioinformatics

Background:

• Computational models of gene regulation are crucial for understanding biological mechanisms in fields like biotechnology and medicine.
• Existing tools for reconstructing whole-genome gene regulatory networks (GRNs) are limited, hindering static and dynamic analyses.
• There is a need for sophisticated, open-source tools to infer comprehensive GRNs from high-throughput gene expression data.

Purpose of the Study:

• To introduce Augusta, an open-source Python package designed for the inference of Gene Regulatory Networks (GRNs) and Boolean Networks (BNs).
• To enable the reconstruction of genome-wide models from high-throughput gene expression data suitable for static and dynamic analyses.
• To provide a user-friendly, command-line operated tool applicable to both model and non-model organisms.

Main Methods:

• Augusta infers GRNs from gene expression data, refining them by predicting transcription factor binding motifs and integrating database interactions.
• The inferred GRN is converted into a draft BN by querying a model database and applying logical rules to gene interactions.
• Models are provided in the Systems Biology Markup Language (SBML) format, allowing for manual editing and further analysis.

Main Results:

• Augusta successfully reconstructs genome-wide GRNs and BNs from gene expression data.
• The package refines network inference by incorporating transcription factor binding site predictions and curated interaction databases.
• The approach is effective even for organisms with limited genomic information available in public databases.
• Augusta facilitates automated model prediction for various genomes via its command-line interface.

Conclusions:

• Augusta offers a robust and versatile solution for inferring gene regulatory and Boolean networks from high-throughput expression data.
• The package enhances the study of gene regulation by enabling static and dynamic analyses of genome-wide models.
• Augusta's applicability to non-model organisms, including microbes, expands the scope of computational systems biology research.
• The open-source nature and SBML compatibility of Augusta promote accessibility and further development in the field.